Keyword Search Terms: How to adequately argue for alternative search parameters

Within early stages of litigation, to successfully negotiate and object to discovery parameters, such as keyword search terms, parties must have a legal justification AND provide adequate evidence to support their objection. Unsupported statements and concerns about producing “an unreasonable number of irrelevant results," or that discovery parameters are not proportional to the matter are only conclusory statements.

The errors made in Custom Hardware Eng'g & Consulting v. Dowell (E.D. Mo. Jan. 3, 2012) showcases how courts are rejecting discovery objections when attorneys don't take the time or know how to support their claims. In this trade secret case, the U.S. District Court rejected the defendant's (Dowell), objections to the plaintiff's (Custom Hardware) production request and additionally ruled that the defendants' alternative list of search terms was invalid because it “would prevent Plaintiff from obtaining discoverable information and is inconsistent with the broad scope of discovery established by Rule 26(b)."

Because of the advanced technical nature of this case, the court ordered the discovery of the defendant's electronically stored information (ESI) to be conducted in a three-step process set forth in Ameriwood Industries, Inc. v. Liberman (E.D. Mo. Dec. 27, 2006):

1. Parties choose a forensic expert to produce forensic copies of defendants' computers
2. Expert recovers ESI and provides data in a searchable form to defendants' counsel
3. Within 20 days of receipt, defendants' counsel must review information for responsiveness and privilege and produce to plaintiff all responsive and non-privileged information

The dispute between the parties occurred in Step 3. After 20 days and receiving Plaintiff's motion to compel production, the defendant refused to comply based on the concern that the plaintiff's search terms were “overbroad." The defendant subsequently supplied their own alternative search terms.

The court rejected all of the defendant's claims, which are outlined below:

• Search terms contained “common terms" producing an “unreasonable number of irrelevant results." The defendant failed to provide the court with (1) any evidence to support this claim and (2) no authority to support this objection. Even if the court was to assume that defendants' intended to make a proportionality claim under Rule 26(b)(2)(C)(iii), the facts would not support this argument. (Based on the conditions provided in the Ameriwood process, the court assumed that searching through the recovered data would be like “conducting a search of a computer hard drive or of a word processing document," presenting a low burden to the defendants.)
• Search terms would have produced privileged documents. Fear of disclosing privileged material gives a defendant no ground to object to a plaintiff's search terms . The defendant had the option exclude all privileged documents without revealing protected information by creating a privilege log.
• Some search terms would have only identified irrelevant information. No evidence was provided by the defendant to support this claim. The court concluded that the plaintiff's search terms may have produced “matter that is relevant to any party's claim or defense."
• The defendant's alternative search terms were “problematic and inappropriate." Rule 26(b)(1) enables parties to conduct discovery with a broach scope . The court ruled that defendant's keyword search terms would violate Rule 26 based on Defendants' search terms requiring an exact match “between the ESI and a keyword search term's phrasing, capitalization, or both," meaning that variations of a word or misspellings would be excluded.

All of the defendant's objections and alternative search terms stemmed from the contention that the plaintiff's keyword search terms were over inclusive of information for production. The court addressed these concerns about keyword searching by citing the Sedona Conference. “Keyword searches end up being both over- and under-inclusive in light of the inherent malleability and ambiguity of spoken and written English." The benefit of keyword searching is only realized when a specified keyword matches a “sometimes arbitrary choice of language to describe the targeted topic of interest," thus adopting the defendant's search terms would have been inappropriate because of its narrow scope in nature.

Practical E-Discovery Advice

In cases like this, parties should always analyze and review how search terms affect potential costs, however parties cannot make empty claims to modify keyword search terms. If parties are going to involve the court, they must be equipped to support their claim with proof that search terms are unreasonable. To prove that search terms are unreasonable parties should utilize a consultant or e-discovery analysis software to establish why “some search terms only identify irrelevant information," and how “the search terms obligate a party to preserve and collect a disproportionate amount of data in relation to the amount in controversy in the case."

Metrics that can project the potential costs of preserving, reviewing and ultimately producing ESI as a result of a party's search terms will allow the objecting party to provide the necessary, documented evidence courts are seeking in deciding proportionality or overly burdensome claims. In this case, the defendant failed to provide the court any reason to grant its request, leading to the unfavorable ruling outlined above.

For more information on recent e-discovery case law and current rulings, view the Exterro-hosted webcast, “2012 E-Discovery Case Law Forecast: Hindsight is 20/20," here.

Mike Hamilton, J.D. is a Sr. E-Discovery Analyst at Exterro, Inc., focusing on educating Exterro customers, prospects and industry experts on how to solve e-discovery issues proactively with technology. His e-discovery knowledge, legal acumen and practical experience give him a valuable perspective on bridging the gap between IT and legal teams. You can find him on Google+, Twitter and Linkedin.